noise problem with TDA1541

[vincent_brient]

I made a no-oversampling DAC using a CS8412 and a TDA1541 (not TDA1541A) chip. The connection between CS8412 and TDA1541 is similar to:
http://users.verat.net/~rogic/1541/1541a_rev1_1b_sch.pdf
I used a NE5532 I/V.
I works fine for a 0dB 1KHz test tone. Could you confirm that the output signal has a 3.7V offset?
For low amplitude signal test I get a very distorded sound, even a -90dB 1KHz test tone make a huge distortion-noise and when I press pause on the CD player the noise totally disappears.

Any idea?

I have succeded to make a NOS dac using AD1862 and an OS dac using AD1955 and CS8412 and I had no problem. I will also test pcm56, ad1865 and pcm 1852.

 

[vincent_brient]

I found th problem. When I solder a 470p capacitor between pin 16 and 17 the problem disappears so my chips are TDA1541A although I read TDA1541 R1 on top of them.

 

[JohnW]

Vincent,

Now it appears you have it working, what does it sound like?

John

 

[Fin]

I found th problem. When I solder a 470p capacitor between pin 16 and 17 the problem disappears so my chips are TDA1541A although I read TDA1541 R1 on top of them.

They might still be the "Non A" version. I have read that Philips may not have included the internal capacitor between pins 16 and 17 - even though it shows it on the datasheet.

How have you dealt with pins 2 and 4?

 

[vincent_brient]

Vincent,

Now it appears you have it working, what does it sound like?

John

I listend the TDA1541 in a non-oversampling configuration, the CD drive was a CD723.
The sound is quite good but not has good as CD723 analog output. CD723 cd player uses a TDA1445 with a 4x IIR (not FIR according to datasheet) and a TDA1445.
Then I modified the CD723 to use the TDA1545A in a non oversampling configutation (using a CS8412) and again the sound was not as good.
I will investigate a will reclock the CS8412 using the CD player clock and if it is still not as good as the original CD723 than I will conclude that the 4x oversampling done in the CD723 is really better than the non-oversampling configuration.
I nice idea would be to connect a TDA1541 and other good DAC after the 4x oversampling.

 

[vincent_brient]

They might still be the "Non A" version. I have read that Philips may not have included the internal capacitor between pins 16 and 17 - even though it shows it on the datasheet.

How have you dealt with pins 2 and 4?

Yes pin 2 and 4 are connected.

 

[Fin]

If they are "Non A" - you can separate them.

 

[guido]

Think it's the other way around

 

[Fin]

Hi Guido

I'm referring to the idea of using pin 2 for BCK - and pin 4 for an input directly from the masterclock.

 

[guido]

Yes, i'm mixed up. My bad.

 

[Fin]

Hi Guido

Do you have a recommendation for the best type of capacitor to put on pins 16 and 17?

The exixting one in my CDP is 680pF polypropylene - generic. Thinking of using something better - maybe Wima........but which one?

 

[jean-paul]

I am not Guido but the polypropylene capacitor is just fine. Only copper foil Styroflex could be somewhat better ( if you can find them ).

Wima is a brand that manufactures MKT, MKP , MKC etc. It is not a quality norm perse although they are good.

 

[Fin]

I am not Guido but the polypropylene capacitor is just fine. Only copper foil Styroflex could be somewhat better ( if you can find them ).

Wima is a brand that manufactures MKT, MKP , MKC etc. It is not a quality norm perse although they are good.

Hi Jean-Paul

You probably guessed that I'm thinking about my Delta 70.2 - and you know better than most what is in there.

I had a closer look at this 680pF cap and it looks like it could be a Philips/BC Components. The same type is also used in the discrete regulator and I/V. I found some Wima FKM2 in the right value and size. Do you know if these are any good?

In another thread, Farguar suggests changing the 220nF generic (they look like Evox Rifa) polyester caps on both sides of the dac for Wimas. Would Wima MKS2 220nF be a worthwhile upgrade?

In yet another thread I read that you don't like Arcam's idea of using 0.47uF tantalums for 4 of the 14 caps on on both sides of the dac. Should they also be replaced with 220nF Wimas?

 

[guido]

Hi Guido

Do you have a recommendation for the best type of capacitor to put on pins 16 and 17?

The exixting one in my CDP is 680pF polypropylene - generic. Thinking of using something better - maybe Wima........but which one?

The best cap is no cap With that i mean have a try at DEM reclocking stuff.

 

[Fin]

The best cap is no cap With that i mean have a try at DEM reclocking stuff.

Hi Guido

I like the idea of DEM reclocking and will hopefully get around to trying some day.

Have you tried it yet?

 

[vincent_brient]

Then I modified the CD723 to use the TDA1545A in a non oversampling configutation (using a CS8412) and again the sound was not as good.
I will investigate a will reclock the CS8412 using the CD player clock and if it is still not as good as the original CD723 than I will conclude that the 4x oversampling done in the CD723 is really better than the non-oversampling configuration.
I nice idea would be to connect a TDA1541 and other good DAC after the 4x oversampling.

I reclocked the digital data after the CS8412 and now it is really better. Now I can say that the no-oversampling configuration sound better than CD723 4x coniguration on my system.

 

[guido]

Hi Guido

I like the idea of DEM reclocking and will hopefully get around to trying some day.

Have you tried it yet?

No, i did put it in a new dac design i'm working on. But i am putting everything in there at once, so it takes too much time to design/build. And i got destracted by the cd80 restauration.

I think i'll have a go at the cd80 first, doing the usuall things and also use the arcam's clock schema to put in a tent clock. And while am at it